A carregar...
Publicação
Target tracking control system for multi-UAV maritime applications
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 803.67 KB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
The importance of Unmanned Aerial Vehicles (UAV) applications is becoming increasingly clear with every new year. While a lot of research has been done in the past, there are still several research areas, topics and applications to unveil and explore. The Air Force Academy Research Center (CIAFA) is working on a new modular control architecture for UAVs that allows for the reutilization of several modules such as trajectory control, computer vision, automatic learning and sensory data fusion, applied to a large spectrum of
missions, which includes automatic landing or terrestrial target tracking based on computer vision. This thesis contributes to the formal implementation of a command and control architecture, to be used in search and target tracking missions in a maritime environment.
The architecture was idealized in Santos (2021), using three main operating modes: search mode, target tracking mode and collision avoidance mode. Specifically, this dissertation contributes to the implementation of the target tracking mode of that architecture, in the same application scenario, considering that after detecting the target, this should be followed in a collaborative way. The main focus of this target tracking mode implementation is a controller that allows for a group of aircrafts to follow a circular path centered at the target coordinates and moves together with it, using the Moving Path Following (MPF) method (Oliveira, Aguiar, and Encarnação (2016) and Jain (2019)). In a first phase, the proposed control architecture is validated through numerical simulations, considering kinematic unicycle type model for the aircrafts and a simulated target. Next, a more realistic approach is adopted, with software in the loop simulations, using an open source model for the considered aircrafts dynamics. In this phase, specific software modules are developed for the considered control system, using an architecture based on Python, ROS, PX4 and Gazebo. Finally, after the implementation and validation of these simulations, a new feature is added: a target detection algorithm based on the received data from one of the UAVs onboard cameras to estimate the target’s position and velocity, using an automatic detection module developed in Alves, Oliveira, Cruz, and Silva (2022).
These simulation closes the implemented multi-UAV control loop system using a computer vision algorithm. The obtained results show the efficiency of the proposed method in a realistic computational simulation context, which allows for a quick transition to eventual flight tests.
Descrição
Palavras-chave
Flight coordination Moving path following Non-linear control UAVs UAV Controlo não linear Coordenação de voo Seguimento de caminho em movimento
Contexto Educativo
Citação
Felix, M. G. (2022). Target tracking control system for multi-UAV maritime applications. Manuscript to be published.
Editora
Academia da Força Aérea